Complaint Handling:Using Effective Customer Service to Improve Business Results

Challenges with Customer Support

The number of incidents reported is increasing and self-service success is decreasing.

According to AFSMI/SSPA benchmark numbers, between 2000 and 2008, the average number of new support cases customer service centers received grew from 6,000 per month to over 40,000 per month, with customers continuing to prefer support via telephone to web or email channels. In addition, the percent of customers who were able to resolve their issues through self-service has decreased from 48% in 2003 to 40% in 2008.

Support’s Perfect Storm Rages On: Incident Volumes Up; Self-Service Success Down, Ragsdale’s Eye On Service, September 25, 2008

How does your organization view customer complaints? It is unlikely that many organizations welcome customer complaints, but, if handled well, customer complaints can be a significant source of information to control costs and actually enhance customer satisfaction. Yet many organizations seem to have the perspective that addressing customers’ concerns is inherently an impediment to maintaining efficient operations. Much of the management effort directed toward customer service reflects this tension. Often, the emphasis is on taking action—any action—to reduce attention from customers and the press. Too much management time and attention spent on this approach to customer service is expensive and wasteful.

Resource-efficiency and customer service are not opposing forces. In fact, using systematic techniques to escalate and investigate customer complaints, take appropriate containment actions, and then choose and implement corrective and preventive action can help most organizations improve customer satisfaction, cut costs, and achieve superior business results.

Systematic Data Collection

What value does your customer support function provide to your organization and your customers?

One measure of their value is the number (or percentage) of customer issues they are able to handle directly and close out during the initial call. Another measure of their value could be the confidence they provide customers that the issues not resolved during the initial call will be handled promptly and effectively. A third measure of their value could be the quantity of information they collect during the initial call that contributes to final resolution of the issue and minimizes the effort required by second and third level support staff to resolve the issue.

Ask yourself these questions:

KT has designed a proactive model for maximizing the value that the customer support function can provide to both your organization and your customers. At its core is a systematic approach for managing the information that customers provide to ensure that issues are properly escalated and problems are properly identified and promptly resolved so that customer expectations are, at minimum, met (or preferably, exceeded).

The Logic of the Model

Customers have higher service expectations and switch if expectations are not met.

According to the Accenture 2007 Global Customer Satisfaction Survey, 33 percent of the respondents in their sample have higher expectations for customer service today than one year ago and 52 percent have higher expectations than five years ago. “For most, ‘better’ means a representative taking more time to answer their questions and a faster response to their inquiries, followed closely by live access to a service representative.

Service is the primary reason for consumers switching providers. According to the same survey, service is more important than price by 20 percentage points as the cause for switching. This is compounded by the fact that not very many consumers (only 5%) consider the service they receive to be excellent, and 41% of the respondents described the quality of service they received as fair, poor or terrible. The ultimate business impact is that “Most respondents (59%) report having switched at least one provider in the past year due to poor service.

Accenture 2007 Global Customer Satisfaction Survey

Most problem resolution systems include the following steps:

A systematic process for gathering the information required to complete each of these steps effectively is at the core of our model for maximizing customer satisfaction and efficiency. Using “process” thinking helps organizations link customer complaint systems to operations and make the leap from customer service to business results.

In working with one of our clients, we established a systematic approach to documenting complaints, the average time to process a ticket dropped to below 5 minutes from almost 10. More importantly the variation in the time required to process a ticket also decreased dramatically, giving management much better predictability and control of their process. Anecdotally, the second line support staff also saw an improvement in the quality of information they were receiving and needed in order to do their work.

In this paper we will discuss how customer service representatives using a structured approach to gather information can enhance an organization’s Issue Escalation, Issue Containment, and Issue Investigation processes.

Issue Identification

To meet service expectations providers employ 19.5 TSEs per customer.

The question of how much companies should spend to support their customer’s demands for high quality service is difficult to answer. In part this is because the demand for service is affected by the complexity of the products and technologies supported, the volume of end-users requiring service (consumer v. B2B), and the size of the company, with smaller firms needing more TSEs per client to provide support. “For support organizations with 100 TSEs or more, the average ratio was one TSE for each 17.2 customers. For support organizations with less than 100 TSEs, the average ratio was one TSE for every 20.4 customers. This data highlights a key challenge for smaller technology companies: Doing more with less is an ongoing refrain.

Member Collaboration: Support Ratios—TSEs to Customers, John Ragsdale, Vice President of Technology

Issue Identification begins with the first phone call, e-mail, or fax received regarding a complaint. Customers may initially provide information that is biased by assumptions or premature and inaccurate conclusions about potential causes. The person handling the initial call needs to gather data that provides a robust understanding of the presenting event and how it relates to the product or service in question. It is probably premature to begin exploring the Why, so the initial questions should focus on the What, Where, When, and Extent of the issue. These questions will help us establish what we know to be true about the presenting issue and how it is an issue, rather than what we think might be the issue. If the answers to these are sufficient, they can guide escalation of the issue, decisions on containment, and provide the starting point for determining the cause of the concern. Accurately describing the issue is crucial to the process. “Device is dead” or “Device doesn’t work” does not add much to our state of knowledge.

When Customer Service representatives collect information about complaints:

Problem Analysis: Finding the Root Cause of Deviations. “A problem well defined is a problem half-solved.” – John Dewey. To facilitate tracking and trending, customer support personnel should use a consistent framework to document information about customer complaints. Kepner-Tregoe’s Problem Analysis provides such framework for clarifying and documenting non-conformances accurately and concisely.

Indications of Need: Handling Customer Complaints

  1. Is it ever unclear whether a complaint is “real” or not?
  2. Are complaints ever ignored or swept under the rug?
  3. Do the same complaints come up again and again?
  4. Is it often unclear whether a complaint should be escalated or not?
  5. Does the same complaint often surface in more than one location or on more than one product or service?
  6. Do you ever get complaints about how you handle complaints?
  7. Are the sources of complaints often isolated from the process of handling complaints?
  8. Is it clear to everyone how seriously a complaint should be regarded?
  9. Are complaints ever delayed or allowed to grow larger and/or get worse?
  10. Do complaints ever turn into public relations problems?

Answer the 10 questions above. If you answer “YES” to 5 or more of the questions, you may have a need for help in your customer complaint handling process.

Case Study Example

This case study provides an example of how an organization with an effective Customer Service data collection protocol was able to promptly recognize a significant issue.

At 9:05 AM this morning, Customer Support received the first call regarding the failure of one of our blood pressure monitors, the BP200 Digital Blood Pressure Monitor. The complaints center on the Liquid Crystal Display which shows the results of the blood pressure reading. Customers are reporting that, at first use, the LCD never comes on, no matter what they do. We have not gotten any calls about our BP300 Digital Blood Pressure Monitor or the BP1000 Wrist Blood Pressure Monitor.

We began to get a few other calls with the same symptoms through the morning. We alerted the regional Help Desks in Seattle, Sacramento, Des Moines, and Montgomery, but none of them had seen anything like this yet.

By gathering information about where the problem was seen “geographically,” we were able see the pattern that most lived in New England. Three did not, but when asked where purchased, we learned the three “outsiders” had bought their BP200s in New England while on vacation.

By the end of the morning we had received 47 calls. We used this information to create what we thought was a good initial description of the event. Because people used a common convention to document these calls, we were able to determine that there was a pattern and that these were probably not isolated events. Using the problem analysis framework, the team summarized the data as shown in Figure 1.

Figure 1:
Customer Service presents initial facts about BP200 LCDs not displaying


What BP200 Digital Blood Pressure Monitor
LCD not displaying - dead
No complaints logged for BP300 Digital Blood Pressure Monitor or BP1000 Wrist Blood Pressure Monitor

LCD displaying wrong, inaccurate readings
Where All called in to Northeast only; All bought in Northeast only No complaints logged from Southwest (Sacramento), Northwest (Seattle), Southeast (Montgomery), Central (Des Moines)
When 3/21/01 9:05 AM
Received complaints throughout the morning
During first customer use.
Before 3/21/0 at 9:05 AM
Need More Data
Need More Data
Extent 47 user complaints so far; Need More Data
One failure per Blood Pressure Monitor
Need More Data
  Copyright © Kepner-Tregoe, Inc. 2010 All Rights Reserved  

Issue Escalation and Prioritization

In any business, there are low-level complaints: problems for which you already know the cause; problems for which a fix has already been decided; and problems that are one-time random occurrences of little seriousness. Then there are the high-level complaints, where a key product is experiencing a damaging failure in increasing numbers. It is crucial that the customer service function gather relevant information and communicate it clearly so that management can appropriately escalate the problem and identify the most qualified experts to address the issue. Perhaps just as importantly, we need to determine who we do not need to involve so that they can allocate their time and effort to other important issues.

When people in your organization gather to determine how to address a customer complaint:

Situation Appraisal: A logical framework for identifying and prioritizing issues. This is especially important when addressing customer complaints where the presenting short-term issue may overshadow more serious long-term concerns. Effectively prioritizing issues and carefully allocating scarce resources is one of the most important elements in a well functioning customer service organization. Kepner-Tregoe’s Situation Appraisal provides a logical framework for prioritizing issues and planning involvement to resolve those issues effectively and efficiently.

Case Study Example

Quality reviewed the information provided by customer service and recognized that they needed to initiate an investigation to determine the cause of BP200 LCD’s not displaying. They had received 47 calls in one day about this one model. The success of this model is critical to our strategic plan to expand in the home health care market.

In addition to minimizing the impact of the issue on customers and the organization, they decided to ask for recommendations on how to contain its effects. The investigation team would be given 24 hours. If we could gather enough data in that time we would focus on solving the problem. If not, we would have to consider recalls, swap-outs, coupons, and the like.

Using the situation appraisal framework, Quality summarized the information and documented their thinking, as shown in Figure 2.

Figure 2:
Management prioritizes BP200 LCDs not displaying

Concerns Current Impact Future Impact Time Frame Plan Next Steps Plan Involvement
Investigate BP200 LCDs not displaying 47 complaints from direct end-users

Northeast only
Success is critical to our strategic plan 30 days to complete Full investigation using Problem Analysis BP200 Product Development to lead

Quality to review
Contain BP200 LCDs not displaying 47 complaints from direct end-users

Northeast only
Success is critical to our strategic plan 3 days to complete Mediate risk using Potential Problem Analysis Marketing to lead

Legal to review

 Copyright © Kepner-Tregoe, Inc. 2010 All Rights Reserved

Issue Containment

Effective containment requires interim actions and contingent actions to limit the spread of any potential problems associated with the issue. Typically, containment actions are taken before you fully understand the cause. It is important to have accurate documentation of the issue so you can recognize what you don’t know, and consider if what you do know is enough to take action. The data gathered by the customer service function should drive your consideration of possible containment solutions.

Also, when responding to a major customer, you are not selecting a single action but a “course of actions.” Your first response, second response, and third response should fit together into a strategy, and should be triggered by specific trends or thresholds in the data. Consider whether each containment action is reversible, whether it can be over-ridden, and how. Do they build seamlessly, or do they make you appear as if you are thrashing around, trying things out?

When people in your organization initially react to a customer complaint:

Potential Problem Analysis: Avoiding Risks Inherent in Actions. Potential Problem Analysis helps people efficiently identify measured responses to a customer complaint, and any potential problems or effects that might result. Preventive and contingent actions are established to minimize risks, while promoting and capitalizing actions are established to extend benefits.

Case Study Example

The team decided that, since there were no reports or evidence of injury and the number of cases was still small, there was no immediate need for a re-call or other drastic actions. They did consider how to best manage product already in the field that might begin to display symptoms and how to manage the concerns of customers already experiencing difficulties.

So the team used the logic of potential problem analysis to implement preventive actions to minimize the probability of customers receiving defective units and planned contingent actions to minimize the seriousness if it did happen, as shown in Figure 3.

Figure 3:
Marketing contains BP200 LCDs not displaying

Potential Problems Likely Causes Preventive Actions Contingent Actions Triggers
Number of defective units continues to grow Distributors continue to sell defective units Fred Fenster of R&D will write up a protocol of a simple test the retailers could do to test for this problem in their stock. Due today Laura Hearne of Communications will write some press releases and liaise with Investor Relations and Legal on the language. Due today

Ron Lin from the factory will check on inventory levels and assess how long it will take to either fly in 1,000 units or ship out 50,000 units. Due this evening
Issue statements if more than 1% of units found to be defective

Initiate recall if more than 5% of units found to be defective

Copyright © Kepner-Tregoe, Inc. 2010 All Rights Reserved

Issue Investigation

Finding true cause usually consumes the time and effort of Technical Support Engineers and other Subject Matter Experts. Ideally, when they begin working on the issue, there is a solid baseline of data that can direct them as they switch from passive data collecting to active information seeking. This can include calling key customers and having them test or examine products, having the factory check their production records or QA logs, or having distributors dig through their records. Make sure you have a plan for gathering whatever data is missing, and a way of keeping the data you are working on up-to-date.

When people in your organization investigate issues:

Problem Analysis: Finding the Root Cause of Deviations. When something goes wrong, the question is, why? Problem analysis provides a powerful logic for understanding the root cause of performance deviations. Corrective action that drives business value rests upon a foundation comprised of an effective, systematic logic for finding root cause.

Case Study Example

A number of hypotheses had been generated about the cause of this problem, some based on fact, others on rumor. Certain managers were particularly interested in looking into the new supplier of the LCDs.

The team used the information gathered during the issue identification step to focus their search for additional information and identify relevant changes. Rather than looking for global changes, the team recognized that whatever caused the BP200 LCD not displaying is only affecting BP200s and only began affecting machines on 21 March.

Using this approach they identified 1) the new battery compartment design, and 2) new cellular assembly process as relevant changes, as shown in Figure 4.

Figure 4:
Investigation Team gathers additional relevant facts

  IS IS NOT Distrinctions Changes
What BP200 Digital Blood Pressure Monitor

LCD not displaying - dead
BP300 Digital Blood Pressure Monitor BP1000 Wrist Blood Pressure Monitor

LCD displaying wrong, inaccurate readings
Uses Batteries

Uses 2 AA batteries in compartment, sideways, versus 3 AAAs in same compartment, lengthwise
Design change - went to one common battery compartment with different battery orientations to save on molding costs - 2/14
Where All called in to Northeast only; All bought in Northeast only No complaints logged from Southwest (Sacramento), Northwest (Seattle), Southeast (Montgomery), Central (Des Moines) All made at the Northborough Plant. New cellular assembly process - multiple models share common steps
When 3/21/01 9:05 AM

During first customer use.
Before 3/21/01 at 9:05 AM

During in-process inspection, during final QA inspection

47 user complaints so far; Need More Data

One failure per Blood Pressure Monitor

Need More Data    

Copyright © Kepner-Tregoe, Inc. 2010 All Rights Reserved

The team used this information to develop the hypothesis that employees in the Northborough Plant, now using a cellular manufacturing process where each cell shares all products, are mixing up the battery orientation and battery type between the BP200 with the BP1000. When they use 3 AAA batteries in the lengthwise orientation, the unit will not make contact with the battery compartment and the LCDs will not light.

The team then tested the original hypothesis and the one they developed against the facts they had gathered to assess how well the hypothesis explained all those facts. The original hypothesis about the new vendor could not explain all the facts and was eliminated from consideration.

Employees installing the battery compartment with three AAA batteries lengthwise best explained the facts. They also identified assumptions required to explain the cause and additional information needed to confirm true cause, as shown in Figure 5.

Figure 5:
Investigation Team tests each hypothesis and documents findings

Possible Cause Hypothesis Does Not Explain Explains Only If
LCDs from new vendor faulty Why not BP1000 (same LCD)?

Received after affected units shipped
Employees installing compartment with AAA batteries lengthwise   If common battery compartment is used in affected units

If QA power test uses AC-to-DC converted power source to test the unit, so as not to drain the batteries

Copyright Kepner-Tregoe, Inc. All rights reserved.

To test this conclusion, the team created a plan to check the returns, and also do some sampling at the stores, to look at the battery orientation and battery type. If the defective units have AAA batteries instead of AA batteries, the cause is confirmed.

These simple checks enabled the manufacturer to find true cause of the customer complaint. The rigorous and timely application of Problem Analysis to the existing data prevented challenging the new vendor, unnecessarily, or the adopting of other expensive alternative courses of action. Redesigning the layout of component parts in the new cellular production area and changing testing protocols prevented recurrence of the problem, improving both productivity and customer satisfaction.


‘When the going gets tough, the tough take really good notes.’ Every time your customers contact you, there is an opportunity to both satisfy their needs and collect valuable information. The initial contact presents an opportunity that may not be available in subsequent communications if customers are not able to remember important details.

How well is your organization using the initial interaction the customer support function has with the consumer to collect information to:

The Kepner-Tregoe processes provide a simple framework for integrating the efforts and information gathered by the customer support function into operations, making customer support compliance part of operations, the basis for continuous improvement, and a catalyst for operations excellence.


Printed From: